| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 200.0, | |
| "global_step": 51600, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 25.0, | |
| "learning_rate": 9.47515299877601e-05, | |
| "loss": 26.7169, | |
| "step": 6450 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "eval_cer": 1.0, | |
| "eval_loss": 3.334904909133911, | |
| "eval_runtime": 8.0603, | |
| "eval_samples_per_second": 26.054, | |
| "eval_steps_per_second": 3.35, | |
| "eval_wer": 1.0, | |
| "step": 6450 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "learning_rate": 8.422521419828641e-05, | |
| "loss": 3.1417, | |
| "step": 12900 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "eval_cer": 0.9991341991341991, | |
| "eval_loss": 3.1051225662231445, | |
| "eval_runtime": 7.9823, | |
| "eval_samples_per_second": 26.308, | |
| "eval_steps_per_second": 3.382, | |
| "eval_wer": 1.0, | |
| "step": 12900 | |
| }, | |
| { | |
| "epoch": 75.0, | |
| "learning_rate": 7.370216238270095e-05, | |
| "loss": 2.8102, | |
| "step": 19350 | |
| }, | |
| { | |
| "epoch": 75.0, | |
| "eval_cer": 0.9298701298701298, | |
| "eval_loss": 2.55711030960083, | |
| "eval_runtime": 8.1516, | |
| "eval_samples_per_second": 25.762, | |
| "eval_steps_per_second": 3.312, | |
| "eval_wer": 1.0, | |
| "step": 19350 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "learning_rate": 6.318074255405957e-05, | |
| "loss": 2.5334, | |
| "step": 25800 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "eval_cer": 0.8883116883116883, | |
| "eval_loss": 2.404860496520996, | |
| "eval_runtime": 8.0509, | |
| "eval_samples_per_second": 26.084, | |
| "eval_steps_per_second": 3.354, | |
| "eval_wer": 1.0, | |
| "step": 25800 | |
| }, | |
| { | |
| "epoch": 125.0, | |
| "learning_rate": 5.2657690738474096e-05, | |
| "loss": 2.4155, | |
| "step": 32250 | |
| }, | |
| { | |
| "epoch": 125.0, | |
| "eval_cer": 0.8571428571428571, | |
| "eval_loss": 2.3261098861694336, | |
| "eval_runtime": 8.0386, | |
| "eval_samples_per_second": 26.124, | |
| "eval_steps_per_second": 3.359, | |
| "eval_wer": 0.9952380952380953, | |
| "step": 32250 | |
| }, | |
| { | |
| "epoch": 150.0, | |
| "learning_rate": 4.213463892288862e-05, | |
| "loss": 2.3529, | |
| "step": 38700 | |
| }, | |
| { | |
| "epoch": 150.0, | |
| "eval_cer": 0.8545454545454545, | |
| "eval_loss": 2.290559768676758, | |
| "eval_runtime": 8.0265, | |
| "eval_samples_per_second": 26.163, | |
| "eval_steps_per_second": 3.364, | |
| "eval_wer": 0.9952380952380953, | |
| "step": 38700 | |
| }, | |
| { | |
| "epoch": 175.0, | |
| "learning_rate": 3.161158710730314e-05, | |
| "loss": 2.3158, | |
| "step": 45150 | |
| }, | |
| { | |
| "epoch": 175.0, | |
| "eval_cer": 0.8363636363636363, | |
| "eval_loss": 2.258056402206421, | |
| "eval_runtime": 8.0276, | |
| "eval_samples_per_second": 26.16, | |
| "eval_steps_per_second": 3.363, | |
| "eval_wer": 0.9952380952380953, | |
| "step": 45150 | |
| }, | |
| { | |
| "epoch": 200.0, | |
| "learning_rate": 2.109016727866177e-05, | |
| "loss": 2.2885, | |
| "step": 51600 | |
| }, | |
| { | |
| "epoch": 200.0, | |
| "eval_cer": 0.8337662337662337, | |
| "eval_loss": 2.2457661628723145, | |
| "eval_runtime": 8.1248, | |
| "eval_samples_per_second": 25.847, | |
| "eval_steps_per_second": 3.323, | |
| "eval_wer": 0.9952380952380953, | |
| "step": 51600 | |
| } | |
| ], | |
| "max_steps": 64500, | |
| "num_train_epochs": 250, | |
| "total_flos": 4.958565694536252e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |